'Precision' One-Channel AC Dimmer

P

P. Short

Super Moderator
Staff member
Probably a pointless exercise, but here is a view of a one-channel dimming AC controller that I'm working while at my ex's place 2000 miles from home.

Somewhat inspired by Dirk's OneSticks, it's 0.8" x 2" in size and is a dimming controller combined with an AC SSR. The controller incorporates a dimming curve table from the get-go capable of controlling the on-timing down to about the 4uS resolution (about 8x better than Renard). It has 128 levels of brightness input, received as a single character at 115200 baud, although there are provisions for daisy-chaining several controllers together. For the potential daisy-chain mode it would use a very simple protocol, using just a '0x00' byte as the frame delimiter and allow data byte values of 1-127. The 4-pin connector on the right side of the PCB is for providing data and power to the controller.

I don't know if this is going to go anywhere, considering the rapidly declining interest these days in AC controllers, so maybe this is just something to entertain myself while away from home.
 

Attachments

  • AC_Ctrl1.png
    AC_Ctrl1.png
    22.9 KB · Views: 35
Re: Single channel AC dimmer

P. Short said:
Probably a pointless exercise, but here is a view of a one-channel dimming AC controller that I'm working while at my ex's place 2000 miles from home.

Somewhat inspired by Dirk's OneSticks, it's 0.8" x 2" in size and is a dimming controller combined with an AC SSR. The controller incorporates a dimming curve table from the get-go capable of controlling the on-timing down to about the 4uS resolution (about 8x better than Renard). It has 128 levels of brightness input, received as a single character at 115200 baud, although there are provisions for daisy-chaining several controllers together. For the potential daisy-chain mode it would use a very simple protocol, using just a '0x00' byte as the frame delimiter and allow data byte values of 1-127. The 4-pin connector on the right side of the PCB is for providing data and power to the controller.

I don't know if this is going to go anywhere, considering the rapidly declining interest these days in AC controllers, so maybe this is just something to entertain myself while away from home.
Click to expand...

2000 miles ? You are as bad as the Canadian who said he would walk 500 miles :)

I hope it goes somewhere Phil . AC power will be around forever and interest will grow/return. Versatile small AC controllers in my view are worth their weight in gold.
Since I have been experimenting/learning/researching as much as I can , I have noticed very few compact AC dimmer/controllers available for DIY . The ones that are generally don't
do more than filament or fan .
Then there is the coding of them which is not very flexible (mostly interrupt and Mqtt) . What would be nice to see if possible is a compact AC dimmer/controller that is flexible for programming and scalable amperage wise. 8A -> 15A max variations would be kinda just right I am thinking .

I know you seem to favor the pic , but it would be good if you fabricated something Arduino friendly .

The ESP32 is my got to favorite but the Wemos pro mini is antenna capable which I also like alot .

Always interested in what you are up to with AC .
 
Re: Single channel AC dimmer

P. Short said:
Probably a pointless exercise, but here is a view of a one-channel dimming AC controller that I'm working while at my ex's place 2000 miles from home.

Somewhat inspired by Dirk's OneSticks, it's 0.8" x 2" in size and is a dimming controller combined with an AC SSR. The controller incorporates a dimming curve table from the get-go capable of controlling the on-timing down to about the 4uS resolution (about 8x better than Renard). It has 128 levels of brightness input, received as a single character at 115200 baud, although there are provisions for daisy-chaining several controllers together. For the potential daisy-chain mode it would use a very simple protocol, using just a '0x00' byte as the frame delimiter and allow data byte values of 1-127. The 4-pin connector on the right side of the PCB is for providing data and power to the controller.

I don't know if this is going to go anywhere, considering the rapidly declining interest these days in AC controllers, so maybe this is just something to entertain myself while away from home.
Click to expand...

Interesting, Phil. How are you providing DC for the PIC? Through a cable such as a POE concept?
 
Re: Single channel AC dimmer

Dirk -

It's meant as an alternative to widgets like that RobotDyn device that Angus referenced in one of the first posts in this thread. Rather than being standalone, it would be used in conjunction with another device like an arduino or the Wemo (although I still don't understand the Wemo ecology very well), and would get both it's data and power from that (or whatever it gets its power from).

It was designed during the recovery time from my two-day straight-through solo Minnesota to California road-trip, and may not necessarily make a lot of sense when I'm finished recovering. And it might be viewed as a masochistic exercise in programming due to the low-end nature of the PIC10F320 microcontroller.
 
Re: Single channel AC dimmer

I suppose the obvious question is 'why'? After all, AC dimmer support already exists in the Arduino world, and this device seems to be just an AC SSR plus a PIC microcontroller. Isn't this just duplicating the existing functionality?

In my opinion, no. My understanding of the Arduino options is limited, but I'm guessing that the design is similar to that of the Renard. Basically, the time interval between successive AC line crossings (perhaps excluding some dead-time around the crossings) is divided into either 255 or 99 equal intervals, and the software selects one of those intervals to send a pulse to activate the triac (or scr) in the AC SSR. This works fine (with some minor limitations) for incandescent lights, not so perfectly for LED lights. The drawback with LED lights is that the usual devices only put out light in a fairly narrow range of time in the cycle, typically centered around the peak of the AC waveform. This means that activating the triac/ssr is useful for a reduced number of those aforementioned time intervals corresponding to that range. Someone bright-eyed and bushy-tailed will be happy with that, not having any idea what is missing.

This device that I'm working on does not have the limitation that each time-slot is the same size, or located at a fixed position within the AC cycle. For each channel there is a persistent table in the microcontroller that defines the start time of each time slot in which to activate the triac/ssr. With the PIC that I'm using each timeslot can be placed with 4uS resolution anywhere after the 'previous' time slot (by comparison, the Renard design each timeslot is around 30uS wide (in 60 Hz power regimes), placed in immediate succession. As an example of the 'new' design, to cater to the LED string that I was messing around with, the first timeslot (for the brightest result) is placed at the midpoint of the AC cycle and is quite wide, while the last timeslots (for the dimmest result) are placed around 7 ms into the 8.33 mS AC cycle, and are fairly narrow. This results smooth, linear dimming ramps over the entire 255 brightness level inputs.

I don't know if the Arduino libraries are capable of results like this, but I'm sure that someone will let me know if they can do this.
 
Re: Single channel AC dimmer

Anyway, I'm done hi-jacking Angus' thread (other than responding to questions). I'll start a new thread when the parts and boards arrive sometime next week if interest warrants.
 
Re: Single channel AC dimmer

I agree.
AC lights will always be around.
Please keep up the work!
angus40 said:
2000 miles ? You are as bad as the Canadian who said he would walk 500 miles :)

I hope it goes somewhere Phil . AC power will be around forever and interest will grow/return. Versatile small AC controllers in my view are worth their weight in gold.
Since I have been experimenting/learning/researching as much as I can , I have noticed very few compact AC dimmer/controllers available for DIY . The ones that are generally don't
do more than filament or fan .
Then there is the coding of them which is not very flexible (mostly interrupt and Mqtt) . What would be nice to see if possible is a compact AC dimmer/controller that is flexible for programming and scalable amperage wise. 8A -> 15A max variations would be kinda just right I am thinking .

I know you seem to favor the pic , but it would be good if you fabricated something Arduino friendly .

The ESP32 is my got to favorite but the Wemos pro mini is antenna capable which I also like alot .

Always interested in what you are up to with AC .
Click to expand...
 
You must log in or register to reply here.
Back
Top